[alsa-devel] wm9713 regression on pxa platforms
Hi Wolfson maintainers, hi Mark,
Recently, the wm9713 support broke on my platforms. I'm pretty sure that a previous harmless warning became an error preventing the probe of my driver (sound/soc/pxa/mioa701_wm9713.c) : the error is shown in [1] ("Control not supported").
The warning "ASoC: no dapm match for Mic B Source" has be there as far as I'm aware for ages, and it never prevented the probe before.
Now, if I apply the ugly patch in [2], everything works as expected ... but that's definitely not the right fix.
Would any of you be able to figure out a fix please ?
Cheers.
On Sat, Oct 26, 2019 at 10:06:10PM +0200, Robert Jarzmik wrote:
Hi Wolfson maintainers, hi Mark,
Recently, the wm9713 support broke on my platforms. I'm pretty sure that a previous harmless warning became an error preventing the probe of my driver (sound/soc/pxa/mioa701_wm9713.c) : the error is shown in [1] ("Control not supported").
The warning "ASoC: no dapm match for Mic B Source" has be there as far as I'm aware for ages, and it never prevented the probe before.
Now, if I apply the ugly patch in [2], everything works as expected ... but that's definitely not the right fix.
Would any of you be able to figure out a fix please ?
I suspect the patch that turned this into a fatal error was this one:
daa480bde6b3 ("ASoC: soc-core: tidyup for snd_soc_dapm_add_routes()")
That said the error is clearly valid and the driver should be fixed. The routes make no sense and are clearly wrong. I strongly suspect the routes should look like this:
{"Mic B Pre Amp", NULL, "Mic B Source"}, {"Mic B Source", "MPB", "MIC2B"},
I am travelling this week, and would need to do a bit more datasheet diving to be sure. But be worth checking if that fixes it and I can fire out a real patch start of next week.
Thanks, Charles
[2] Ugly test diff --git a/sound/soc/codecs/wm9713.c b/sound/soc/codecs/wm9713.c index 6497c1ea6228..ec140ccbab0b 100644 --- a/sound/soc/codecs/wm9713.c +++ b/sound/soc/codecs/wm9713.c @@ -630,8 +630,8 @@ static const struct snd_soc_dapm_route wm9713_audio_map[] = { {"Mic A Source", "Mic 1", "MIC1"}, {"Mic A Source", "Mic 2 A", "MIC2A"}, {"Mic A Source", "Mic 2 B", "Mic B Source"},
{"Mic B Pre Amp", "MPB", "Mic B Source"},{"Mic B Source", NULL, "MIC2B"},
/* {"Mic B Pre Amp", "MPB", "Mic B Source"}, *//* {"Mic B Source", NULL, "MIC2B"}, */
Charles Keepax ckeepax@opensource.cirrus.com writes:
{"Mic B Pre Amp", NULL, "Mic B Source"}, {"Mic B Source", "MPB", "MIC2B"},
That fixes it on my side.
I am travelling this week, and would need to do a bit more datasheet diving to be sure. But be worth checking if that fixes it and I can fire out a real patch start of next week.
Ok, once you're back, you can added my : Tested-by: Robert Jarzmik <robert.jarzmik.fr>
Cheers.
-- Robert
participants (2)
-
Charles Keepax -
Robert Jarzmik